---
title: "Unlocking Implicit Experience — paper by Xu et al. (2026)"
description: "Training LLM agents for multi-turn tool use is limited by a shortage of realistic trajectories. 24 references, cited by 4 papers here."
authors: ["Zhihao Xu", "Rumei Li", "Jiahuan Li", "Rongxiang Weng", "Jingang Wang", "Xunliang Cai", "Xiting Wang"]
year: 2026
venue: "Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)"
source: "https://aclanthology.org/2026.acl-long.452/"
doi: "10.18653/v1/2026.acl-long.452"
licence: "CC BY 4.0 (ACL Anthology)"
licence_url: "https://creativecommons.org/licenses/by/4.0/"
summary_licence: "Elamur Cosmic License"
canonical: "https://research.elamur.ai/free/en/papers/xu-2026-unlocking-implicit-experience-synthesizing-tool/"
---
[Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/xu-2026-unlocking-implicit-experience-synthesizing-tool/)

[home](https://md.research.elamur.ai/free/en/) / [papers](https://md.research.elamur.ai/free/en/papers/) / [2026](https://md.research.elamur.ai/free/en/papers/2026/page/4/) / Unlocking Implicit Experience: Synthesizing Tool-Use Trajectories from Text

# Unlocking Implicit Experience: Synthesizing Tool-Use Trajectories from Text

**Authors:** Zhihao Xu, [Rumei Li](https://md.research.elamur.ai/free/en/authors/rumei-li/), [Jiahuan Li](https://md.research.elamur.ai/free/en/authors/jiahuan-li/), Rongxiang Weng, [Jingang Wang](https://md.research.elamur.ai/free/en/authors/jingang-wang/), [Xunliang Cai](https://md.research.elamur.ai/free/en/authors/xunliang-cai/) and [Xiting Wang](https://md.research.elamur.ai/free/en/authors/xiting-wang/) · **Year:** 2026 · **Venue:** [Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)](https://md.research.elamur.ai/free/en/venues/acl/) ([ACL 2026](https://md.research.elamur.ai/free/en/venues/acl/2026/)) · **Publisher:** [Association for Computational Linguistics](https://md.research.elamur.ai/free/en/publishers/association-for-computational-linguistics/) · **Licence:** [CC BY 4.0 (ACL Anthology)](https://creativecommons.org/licenses/by/4.0/)

**Links:** [Publisher's page](https://aclanthology.org/2026.acl-long.452/) · [PDF](https://aclanthology.org/2026.acl-long.452.pdf) · [DOI 10.18653/v1/2026.acl-long.452](https://doi.org/10.18653/v1/2026.acl-long.452) · [OpenAlex](https://openalex.org/W7166806951)

**Citations:** 0 citations (OpenAlex, 2026-10-14) · [4](https://md.research.elamur.ai/free/en/papers/xu-2026-unlocking-implicit-experience-synthesizing-tool/cited-by/) in our library, of the 149,439 papers whose bibliographies we read

## Summary
Training LLM agents for multi-turn tool use is limited by a shortage of realistic trajectories. Most existing synthesis methods start from a fixed API collection, so the data can only be as broad as that collection. The authors invert this approach. They argue that ordinary web text, such as how-to guides and procedural descriptions, already records human multi-step problem solving that can be turned into agent data. Their pipeline, GEM, works in four stages:
1. It keeps text segments that describe multi-step procedures.
2. It derives abstract workflows and OpenAI-format tool definitions from them.
3. It has a strong model ground these into user–assistant–tool dialogues.
4. It increases complexity and diversity, then checks the results with rule-based and LLM-based verification.

To reduce cost, they also fine-tune a Trajectory Synthesizer that performs the whole text-to-trajectory conversion in one step and reportedly matches the pipeline's quality. Models trained on GEM data improve substantially on BFCL V3 Multi-turn and τ²-bench. Their out-of-domain data competes with τ-bench in-domain training, which suggests the text-derived data generalizes well.

## Key points
- **Source corpus analysis:** Of roughly 250,000 sampled Ultra-fineweb segments, about 14% contain explicit multi-step workflows. These cover many categories, including customer support, research and data, and education. Qwen3-8B was used for annotation and filtering.
- **Headline gains:** GEM-32B gains 14.9% on BFCL V3 Multi-turn, and GEM training yields a 13.8% gain on τ²-bench.
- **Ablation on BFCL V3 (overall accuracy):**
  - Qwen3-32B rises from 28.35 to 44.88; without the refinement stage it reaches only 32.50.
  - Qwen3-8B rises from 18.00 to 30.25.
- **Data scale and complexity:** Scaling GEM-8B training data from 10K to 70K trajectories keeps improving τ-bench Airline/Retail scores. The synthesized trajectories are long, averaging about 16 messages each.
- **Limitations:** The framework currently handles English, text-only data. The authors name multilingual and multimodal extensions, and use of the data for mid-training, as future work.

*Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.*

## Institutions

- [Renmin University of China](https://md.research.elamur.ai/free/en/universities/renmin-university-of-china/) — University, Beijing, China · 114 works · 16,118 citations
- [Meituan](https://md.research.elamur.ai/free/en/companies/meituan/) — Company, Beijing, China · 17 works · 3,769 citations

📚 Shelves (5): [🖥️ Computer Science](https://md.research.elamur.ai/free/en/shelves/computer-science/) · 16,667 works · [🤖 Artificial Intelligence](https://md.research.elamur.ai/free/en/shelves/computer-science/artificial-intelligence/) · 16,194 works · [🎭 Arts and Humanities](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/) · 16,039 works · [🗣️ Language and Linguistics](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/language-and-linguistics/) · 15,891 works · [👁️ Computer Vision and Pattern Recognition](https://md.research.elamur.ai/free/en/shelves/computer-science/computer-vision-and-pattern-recognition/) · 2,869 works

## Cite

Cite as: Zhihao Xu, Rumei Li, Jiahuan Li, Rongxiang Weng, Jingang Wang, Xunliang Cai and Xiting Wang. 2026. Unlocking Implicit Experience: Synthesizing Tool-Use Trajectories from Text. Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers).

The paper itself, not this page. BibTeX from the library's record of the work.

```bibtex
@inproceedings{xu2026unlocking,
  title = {{Unlocking Implicit Experience: Synthesizing Tool-Use Trajectories from Text}},
  author = {Zhihao Xu, Rumei Li, Jiahuan Li, Rongxiang Weng, Jingang Wang, Xunliang Cai and Xiting Wang},
  booktitle = {Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)},
  year = {2026},
  doi = {10.18653/v1/2026.acl-long.452},
  url = {https://doi.org/10.18653/v1/2026.acl-long.452}
}
```

- [← previous UniversalRAG: Retrieval-Augmented Generation over Corpora of Diverse Modalities and Granularities](https://md.research.elamur.ai/free/en/papers/yeo-2026-universalrag-retrieval-augmented-generation-over/)
- [next → Unlocking Multilingual Reasoning Capability of LLMs and LVLMs through Representation Engineering](https://md.research.elamur.ai/free/en/papers/li-2026-unlocking-multilingual-reasoning-capability-llms/)
- [↑ back to papers of 2026](https://md.research.elamur.ai/free/en/papers/2026/page/4/)

## References

24 references, 6 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

1. Victor Barres, Honghua Dong, Soham Ray, Xujie Si, and Karthik Narasimhan. 2025. τ 2 -bench: Evaluating conversational agents in a dual-control environment. arXiv preprint arXiv:2506.07982. [find](https://scholar.google.com/scholar?q=Victor+Barres%2C+Honghua+Dong%2C+Soham+Ray%2C+Xujie+Si%2C+and+Karthik+Narasimhan%2E+2025%2E+%CF%84+2+%2Dbench%3A+Evaluating+conversational+agents+in+a+dual%2Dcontrol+environment%2E+arXiv+preprint+arXiv%3A2506%2E07982%2E)
2. Chen Chen, Xinlong Hao, Weiwen Liu, Xu Huang, Xingshan Zeng, Shuai Yu, Dexun Li, Shuai Wang, Weinan Gan, Yuefeng Huang, and 1 others. 2025. Acebench: Who wins the match point in tool usage? arXiv preprint arXiv:2501.12851. — [arXiv:2501.12851](https://arxiv.org/abs/2501.12851) [find](https://scholar.google.com/scholar?q=Chen+Chen%2C+Xinlong+Hao%2C+Weiwen+Liu%2C+Xu+Huang%2C+Xingshan+Zeng%2C+Shuai+Yu%2C+Dexun+Li%2C+Shuai+Wang%2C+Weinan+Gan%2C+Yuefeng+Huang%2C+and+1+others%2E+2025%2E+Acebench%3A+Who+wins+the+match+point+in+tool+usage%3F+arXiv+prep)
3. Zhicheng Guo, Sijie Cheng, Hao Wang, Shihao Liang, Yujia Qin, Peng Li, Zhiyuan Liu, Maosong Sun, and Yang Liu. 2024. Stabletoolbench: Towards stable large-scale benchmarking on tool learning of large language models. arXiv preprint arXiv:2403.07714. [find](https://scholar.google.com/scholar?q=Zhicheng+Guo%2C+Sijie+Cheng%2C+Hao+Wang%2C+Shihao+Liang%2C+Yujia+Qin%2C+Peng+Li%2C+Zhiyuan+Liu%2C+Maosong+Sun%2C+and+Yang+Liu%2E+2024%2E+Stabletoolbench%3A+Towards+stable+large%2Dscale+benchmarking+on+tool+learning+of+large+)
4. Ai Jian, Xiaoyun Zhang, Wanrou Du, Jingqing Ruan, Jiangbo Pei, Weipeng Zhang, Ke Zeng, and Xunliang Cai. 2026. Trust-sql: Tool-integrated multi-turn reinforcement learning for text-to-sql over unknown schemas. arXiv preprint arXiv:2603.16448. — [arXiv:2603.16448](https://arxiv.org/abs/2603.16448) [find](https://scholar.google.com/scholar?q=Ai+Jian%2C+Xiaoyun+Zhang%2C+Wanrou+Du%2C+Jingqing+Ruan%2C+Jiangbo+Pei%2C+Weipeng+Zhang%2C+Ke+Zeng%2C+and+Xunliang+Cai%2E+2026%2E+Trust%2Dsql%3A+Tool%2Dintegrated+multi%2Dturn+reinforcement+learning+for+text%2Dto%2Dsql+over+unknown)
5. Mahnaz Koupaee and William Yang Wang. 2018. Wikihow: A large scale text summarization dataset. arXiv preprint arXiv:1810.09305. [find](https://scholar.google.com/scholar?q=Mahnaz+Koupaee+and+William+Yang+Wang%2E+2018%2E+Wikihow%3A+A+large+scale+text+summarization+dataset%2E+arXiv+preprint+arXiv%3A1810%2E09305%2E)
6. Jian Li, Yizhang Jin, Dongqi Liu, Hang Ding, Jiafu Wu, Dongsheng Chen, Yunhang Shen, Yulei Qin, Ying Tai, Chengjie Wang, and 1 others. 2026. Se-search: Self-evolving search agent via memory and dense reward. arXiv preprint arXiv:2603.03293. — [arXiv:2603.03293](https://arxiv.org/abs/2603.03293) [find](https://scholar.google.com/scholar?q=Jian+Li%2C+Yizhang+Jin%2C+Dongqi+Liu%2C+Hang+Ding%2C+Jiafu+Wu%2C+Dongsheng+Chen%2C+Yunhang+Shen%2C+Yulei+Qin%2C+Ying+Tai%2C+Chengjie+Wang%2C+and+1+others%2E+2026%2E+Se%2Dsearch%3A+Self%2Devolving+search+agent+via+memory+and+dense+)
7. Yuetai Li, Huseyin A Inan, Xiang Yue, Wei-Ning Chen, Lukas Wutschitz, Janardhan Kulkarni, Radha Poovendran, Robert Sim, and Saravan Rajmohan. 2025. Simulating environments with reasoning models for agent training. arXiv preprint arXiv:2511.01824. — [arXiv:2511.01824](https://arxiv.org/abs/2511.01824) [find](https://scholar.google.com/scholar?q=Yuetai+Li%2C+Huseyin+A+Inan%2C+Xiang+Yue%2C+Wei%2DNing+Chen%2C+Lukas+Wutschitz%2C+Janardhan+Kulkarni%2C+Radha+Poovendran%2C+Robert+Sim%2C+and+Saravan+Rajmohan%2E+2025%2E+Simulating+environments+with+reasoning+models+for+ag)
8. Aixin Liu, Aoxue Mei, Bangcai Lin, Bing Xue, Bingxuan Wang, Bingzheng Xu, Bochao Wu, Bowei Zhang, Chaofan Lin, Chen Dong, and 1 others. 2025. [Deepseek-v3](https://md.research.elamur.ai/free/en/papers/deepseek-ai-2025-deepseek-v3-2-pushing/). 2: Pushing the frontier of open large language models. arXiv preprint arXiv:2512.02556.
9. Weiwen Liu, Xu Huang, Xingshan Zeng, Xinlong Hao, Shuai Yu, Dexun Li, Shuai Wang, Weinan Gan, Zhengying Liu, Yuanqing Yu, and 1 others. 2024a. Toolace: Winning the points of llm function calling. arXiv preprint arXiv:2409.00920. [find](https://scholar.google.com/scholar?q=Weiwen+Liu%2C+Xu+Huang%2C+Xingshan+Zeng%2C+Xinlong+Hao%2C+Shuai+Yu%2C+Dexun+Li%2C+Shuai+Wang%2C+Weinan+Gan%2C+Zhengying+Liu%2C+Yuanqing+Yu%2C+and+1+others%2E+2024a%2E+Toolace%3A+Winning+the+points+of+llm+function+calling%2E+arXi)
10. Zuxin Liu, Thai Hoang, Jianguo Zhang, Ming Zhu, Tian Lan, Juntao Tan, Weiran Yao, Zhiwei Liu, Yihao Feng, Rithesh RN, and 1 others. 2024b. Apigen: Automated pipeline for generating verifiable and diverse function-calling datasets. Advances in Neural Information Processing Systems, 37:54463–54482. [find](https://scholar.google.com/scholar?q=Zuxin+Liu%2C+Thai+Hoang%2C+Jianguo+Zhang%2C+Ming+Zhu%2C+Tian+Lan%2C+Juntao+Tan%2C+Weiran+Yao%2C+Zhiwei+Liu%2C+Yihao+Feng%2C+Rithesh+RN%2C+and+1+others%2E+2024b%2E+Apigen%3A+Automated+pipeline+for+generating+verifiable+and+dive)
11. Shishir G Patil, Huanzhi Mao, Fanjia Yan, Charlie Cheng-Jie Ji, Vishnu Suresh, Ion Stoica, and Joseph E Gonzalez. 2025. The berkeley function calling leaderboard (bfcl): From tool use to agentic evaluation of large language models. In Forty-second International Conference on Machine Learning. [find](https://scholar.google.com/scholar?q=Shishir+G+Patil%2C+Huanzhi+Mao%2C+Fanjia+Yan%2C+Charlie+Cheng%2DJie+Ji%2C+Vishnu+Suresh%2C+Ion+Stoica%2C+and+Joseph+E+Gonzalez%2E+2025%2E+The+berkeley+function+calling+leaderboard+%28bfcl%29%3A+From+tool+use+to+agentic+evalu)
12. Akshara Prabhakar, Zuxin Liu, Ming Zhu, Jianguo Zhang, Tulika Awalgaonkar, Shiyu Wang, Zhiwei Liu, Haolin Chen, Thai Hoang, Juan Carlos Niebles, and 1 others. 2025. [Apigen-mt: Agentic pipeline for multi-turn data generation via simulated agent-human interplay](https://md.research.elamur.ai/free/en/papers/prabhakar-2025-apigen-mt-agentic-pipeline-multi/). arXiv preprint arXiv:2504.03601.
13. Yujia Qin, Shihao Liang, Yining Ye, Kunlun Zhu, Lan Yan, Yaxi Lu, Yankai Lin, Xin Cong, Xiangru Tang, Bill Qian, and 1 others. 2023. [Toolllm: Facilitating large language models to master 16000+ real-world apis](https://md.research.elamur.ai/free/en/papers/qin-2023-toolllm-facilitating-large-language-models/). arXiv preprint arXiv:2307.16789.
14. Jinbo Su, Lingzhe Gao, Wei Li, Shihao Liu, Haojie Lei, Xinyi Wang, Yuanzhao Guo, Ke Wang, Daiting Shi, and Dawei Yin. 2025. Racqc: Advanced retrieval-augmented generation for chinese query correction. In Findings of the Association for Computational Linguistics: EMNLP 2025, pages 675–689. [find](https://scholar.google.com/scholar?q=Jinbo+Su%2C+Lingzhe+Gao%2C+Wei+Li%2C+Shihao+Liu%2C+Haojie+Lei%2C+Xinyi+Wang%2C+Yuanzhao+Guo%2C+Ke+Wang%2C+Daiting+Shi%2C+and+Dawei+Yin%2E+2025%2E+Racqc%3A+Advanced+retrieval%2Daugmented+generation+for+chinese+query+correction%2E)
15. Shunyu Yao, Noah Shinn, Pedram Razavi, and Karthik Narasimhan. 2024. [τ -bench: A benchmark for tool-agent-user interaction in real-world domains](https://md.research.elamur.ai/free/en/papers/yao-2024-tau-bench-benchmark-tool/). arXiv preprint arXiv:2406.12045.
16. Zhiyuan Yao, Zishan Xu, Yifu Guo, Zhiguang Han, Cheng Yang, Shuo Zhang, Weinan Zhang, Xingshan Zeng, and Weiwen Liu. 2026. Toolace-mcp: Generalizing history-aware routing from mcp tools to the agent web. arXiv preprint arXiv:2601.08276. — [arXiv:2601.08276](https://arxiv.org/abs/2601.08276) [find](https://scholar.google.com/scholar?q=Zhiyuan+Yao%2C+Zishan+Xu%2C+Yifu+Guo%2C+Zhiguang+Han%2C+Cheng+Yang%2C+Shuo+Zhang%2C+Weinan+Zhang%2C+Xingshan+Zeng%2C+and+Weiwen+Liu%2E+2026%2E+Toolace%2Dmcp%3A+Generalizing+history%2Daware+routing+from+mcp+tools+to+the+agent+w)
17. Fan Yin, Zifeng Wang, I-Hung Hsu, Jun Yan, Ke Jiang, Yanfei Chen, Jindong Gu, Long Le, Kai-Wei Chang, Chen-Yu Lee, and 1 others. 2025. Magnet: Multi-turn tool-use data synthesis and distillation via graph translation. In Proceedings of the 63rd Annual Meeting of the Association for Computational Lin… [find](https://scholar.google.com/scholar?q=Fan+Yin%2C+Zifeng+Wang%2C+I%2DHung+Hsu%2C+Jun+Yan%2C+Ke+Jiang%2C+Yanfei+Chen%2C+Jindong+Gu%2C+Long+Le%2C+Kai%2DWei+Chang%2C+Chen%2DYu+Lee%2C+and+1+others%2E+2025%2E+Magnet%3A+Multi%2Dturn+tool%2Duse+data+synthesis+and+distillation+via+g)
18. Xingshan Zeng, Weiwen Liu, Lingzhi Wang, Liangyou Li, Fei Mi, Yasheng Wang, Lifeng Shang, Xin Jiang, and Qun Liu. 2025. Toolace-mt: Non-autoregressive generation for agentic multi-turn interaction. arXiv preprint arXiv:2508.12685. — [arXiv:2508.12685](https://arxiv.org/abs/2508.12685) [find](https://scholar.google.com/scholar?q=Xingshan+Zeng%2C+Weiwen+Liu%2C+Lingzhi+Wang%2C+Liangyou+Li%2C+Fei+Mi%2C+Yasheng+Wang%2C+Lifeng+Shang%2C+Xin+Jiang%2C+and+Qun+Liu%2E+2025%2E+Toolace%2Dmt%3A+Non%2Dautoregressive+generation+for+agentic+multi%2Dturn+interaction%2E+ar)
19. Xiaoyun Zhang, Jingqing Ruan, Xing Ma, Yawen Zhu, Jiansong Chen, Ke Zeng, and Xunliang Cai. 2025. A reasoner for real-world event detection: Scaling reinforcement learning via adaptive perplexity-aware sampling strategy. In Proceedings of the 2025 Conference on Empirical Methods in Natural Language… [find](https://scholar.google.com/scholar?q=Xiaoyun+Zhang%2C+Jingqing+Ruan%2C+Xing+Ma%2C+Yawen+Zhu%2C+Jiansong+Chen%2C+Ke+Zeng%2C+and+Xunliang+Cai%2E+2025%2E+A+reasoner+for+real%2Dworld+event+detection%3A+Scaling+reinforcement+learning+via+adaptive+perplexity%2Dawar)
20. Weikang Zhao, Xili Wang, Chengdi Ma, Lingbin Kong, Zhaohua Yang, Mingxiang Tuo, Xiaowei Shi, Yitao Zhai, and Xunliang Cai. 2025. Mua-rl: Multi-turn user-interacting agent reinforcement learning for agentic tool use. arXiv preprint arXiv:2508.18669. — [arXiv:2508.18669](https://arxiv.org/abs/2508.18669) [find](https://scholar.google.com/scholar?q=Weikang+Zhao%2C+Xili+Wang%2C+Chengdi+Ma%2C+Lingbin+Kong%2C+Zhaohua+Yang%2C+Mingxiang+Tuo%2C+Xiaowei+Shi%2C+Yitao+Zhai%2C+and+Xunliang+Cai%2E+2025%2E+Mua%2Drl%3A+Multi%2Dturn+user%2Dinteracting+agent+reinforcement+learning+for+ag)
21. Yaowei Zheng, Richong Zhang, Junhao Zhang, Yanhan Ye, Zheyan Luo, Zhangchi Feng, and Yongqiang Ma. 2024. [Llamafactory: Unified efficient fine-tuning of 100+ language models](https://md.research.elamur.ai/free/en/papers/zheng-2024-llamafactory-unified-efficient-fine-tuning/). arXiv preprint arXiv:2403.13372.
22. Kimi Team, Yifan Bai, Yiping Bao, Guanduo Chen, Jiahao Chen, Ningxin Chen, Ruijue Chen, Yanru Chen, Yuankun Chen, Yutian Chen, and 1 others. 2025. [Kimi k2: Open agentic intelligence](https://md.research.elamur.ai/free/en/papers/kimi-team-2025-kimi-k2-open-agentic/). arXiv preprint arXiv:2507.20534.
23. Yudong Wang, Zixuan Fu, Jie Cai, Peijun Tang, Hongya Lyu, Yewei Fang, Zhi Zheng, Jie Zhou, Guoyang Zeng, Chaojun Xiao, and 1 others. 2025. Ultrafineweb: Efficient data filtering and verification for high-quality llm training data. arXiv preprint arXiv:2505.05427. — [arXiv:2505.05427](https://arxiv.org/abs/2505.05427) [find](https://scholar.google.com/scholar?q=Yudong+Wang%2C+Zixuan+Fu%2C+Jie+Cai%2C+Peijun+Tang%2C+Hongya+Lyu%2C+Yewei+Fang%2C+Zhi+Zheng%2C+Jie+Zhou%2C+Guoyang+Zeng%2C+Chaojun+Xiao%2C+and+1+others%2E+2025%2E+Ultrafineweb%3A+Efficient+data+filtering+and+verification+for+h)
24. Zhangchen Xu, Adriana Meza Soria, Shawn Tan, Anurag Roy, Ashish Sunil Agrawal, Radha Poovendran, and Rameswar Panda. 2025. Toucan: Synthesizing 1.5 m tool-agentic data from real-world mcp environments. arXiv preprint arXiv:2510.01179. — [arXiv:2510.01179](https://arxiv.org/abs/2510.01179) [find](https://scholar.google.com/scholar?q=Zhangchen+Xu%2C+Adriana+Meza+Soria%2C+Shawn+Tan%2C+Anurag+Roy%2C+Ashish+Sunil+Agrawal%2C+Radha+Poovendran%2C+and+Rameswar+Panda%2E+2025%2E+Toucan%3A+Synthesizing+1%2E5+m+tool%2Dagentic+data+from+real%2Dworld+mcp+environments)

*Updated 31 October 2026*